Closed NeonDaniel closed 7 months ago
Update RMQ config so LLMs have fulls access to queues starting with their name
Observed error logs:
│ 2023-12-06 01:34:01.882 - OVOS - neon_mq_connector.connector:observe_consumers:711 - INFO - Consumer "neon_llm_chat_gpt_score" is dead, restarting │ │ 2023-12-06 01:34:01.905 - OVOS - neon_mq_connector.utils.connection_utils:wrapper:124 - WARNING - ConsumerThread.__init__(args=(), kwargs={'connection_params': <ConnectionParameters host=neon-rabbitmq port=5672 virtual_host=/llm ssl=False>, 'queue': 'chat_gpt_score_input', 'queue_reset': False, 'callba │ │ 2023-12-06 01:34:01.906 - OVOS - neon_mq_connector.utils.connection_utils:wrapper:125 - INFO - Timeout for 5 secs │ │ 2023-12-06 01:34:06.912 - OVOS - neon_mq_connector.utils.connection_utils:wrapper:92 - INFO - Retrying <function ConsumerThread.__init__ at 0x7f0146e11940> execution. Attempt #2 │ │ 2023-12-06 01:34:06.939 - OVOS - neon_mq_connector.utils.connection_utils:wrapper:124 - WARNING - ConsumerThread.__init__(args=(), kwargs={'connection_params': <ConnectionParameters host=neon-rabbitmq port=5672 virtual_host=/llm ssl=False>, 'queue': 'chat_gpt_score_input', 'queue_reset': False, 'callba │ │ 2023-12-06 01:34:06.941 - OVOS - neon_mq_connector.utils.connection_utils:wrapper:125 - INFO - Timeout for 10 secs │ │ 2023-12-06 01:34:16.952 - OVOS - neon_mq_connector.utils.connection_utils:wrapper:92 - INFO - Retrying <function ConsumerThread.__init__ at 0x7f0146e11940> execution. Attempt #3 │ │ 2023-12-06 01:34:16.968 - OVOS - neon_mq_connector.utils.connection_utils:wrapper:124 - WARNING - ConsumerThread.__init__(args=(), kwargs={'connection_params': <ConnectionParameters host=neon-rabbitmq port=5672 virtual_host=/llm ssl=False>, 'queue': 'chat_gpt_score_input', 'queue_reset': False, 'callba │ │ 2023-12-06 01:34:16.969 - OVOS - neon_mq_connector.utils.connection_utils:wrapper:125 - INFO - Timeout for 20 secs │ │ 2023-12-06 01:34:36.990 - OVOS - neon_mq_connector.utils.connection_utils:wrapper:128 - ERROR - Failed to execute after 3 attempts
Description
Update RMQ config so LLMs have fulls access to queues starting with their name
Issues
Other Notes
Observed error logs: